nokogiri 1.8.5 → 1.13.9

data/lib/nokogiri.rb

@@ -1,96 +1,67 @@
-# -*- coding: utf-8 -*-
-# Modify the PATH on windows so that the external DLLs will get loaded.
+# coding: utf-8
+# frozen_string_literal: true
 
-require 'rbconfig'
+require "rbconfig"
 
 if defined?(RUBY_ENGINE) && RUBY_ENGINE == "jruby"
-  # The line below caused a problem on non-GAE rack environment.
-  # unless defined?(JRuby::Rack::VERSION) || defined?(AppEngine::ApiProxy)
-  #
-  # However, simply cutting defined?(JRuby::Rack::VERSION) off resulted in
-  # an unable-to-load-nokogiri problem. Thus, now, Nokogiri checks the presense
-  # of appengine-rack.jar in $LOAD_PATH. If Nokogiri is on GAE, Nokogiri
-  # should skip loading xml jars. This is because those are in WEB-INF/lib and
-  # already set in the classpath.
-  unless $LOAD_PATH.to_s.include?("appengine-rack")
-    require 'stringio'
-    require 'isorelax.jar'
-    require 'jing.jar'
-    require 'nekohtml.jar'
-    require 'nekodtd.jar'
-    require 'xercesImpl.jar'
-    require 'serializer.jar'
-    require 'xalan.jar'
-    require 'xml-apis.jar'
-  end
+  require_relative "nokogiri/jruby/dependencies"
 end
 
-begin
-  RUBY_VERSION =~ /(\d+\.\d+)/
-  require "nokogiri/#{$1}/nokogiri"
-rescue LoadError
-  require 'nokogiri/nokogiri'
-end
-require 'nokogiri/version'
-require 'nokogiri/syntax_error'
-require 'nokogiri/xml'
-require 'nokogiri/xslt'
-require 'nokogiri/html'
-require 'nokogiri/decorators/slop'
-require 'nokogiri/css'
-require 'nokogiri/html/builder'
+require_relative "nokogiri/extension"
 
 # Nokogiri parses and searches XML/HTML very quickly, and also has
 # correctly implemented CSS3 selector support as well as XPath 1.0
 # support.
 #
 # Parsing a document returns either a Nokogiri::XML::Document, or a
-# Nokogiri::HTML::Document depending on the kind of document you parse.
+# Nokogiri::HTML4::Document depending on the kind of document you parse.
 #
 # Here is an example:
 #
-#   require 'nokogiri'
-#   require 'open-uri'
+#     require 'nokogiri'
+#     require 'open-uri'
 #
-#   # Get a Nokogiri::HTML:Document for the page we’re interested in...
+#     # Get a Nokogiri::HTML4::Document for the page we’re interested in...
 #
-#   doc = Nokogiri::HTML(open('http://www.google.com/search?q=tenderlove'))
+#     doc = Nokogiri::HTML4(URI.open('http://www.google.com/search?q=tenderlove'))
 #
-#   # Do funky things with it using Nokogiri::XML::Node methods...
+#     # Do funky things with it using Nokogiri::XML::Node methods...
 #
-#   ####
-#   # Search for nodes by css
-#   doc.css('h3.r a.l').each do |link|
-#     puts link.content
-#   end
+#     ####
+#     # Search for nodes by css
+#     doc.css('h3.r a.l').each do |link|
+#       puts link.content
+#     end
 #
-# See Nokogiri::XML::Searchable#css for more information about CSS searching.
-# See Nokogiri::XML::Searchable#xpath for more information about XPath searching.
+# See also:
+#
+# - Nokogiri::XML::Searchable#css for more information about CSS searching
+# - Nokogiri::XML::Searchable#xpath for more information about XPath searching
 module Nokogiri
   class << self
     ###
     # Parse an HTML or XML document.  +string+ contains the document.
-    def parse string, url = nil, encoding = nil, options = nil
+    def parse(string, url = nil, encoding = nil, options = nil)
       if string.respond_to?(:read) ||
-          /^\s*<(?:!DOCTYPE\s+)?html[\s>]/i === string[0, 512]
+          /^\s*<(?:!DOCTYPE\s+)?html[\s>]/i.match?(string[0, 512])
         # Expect an HTML indicator to appear within the first 512
         # characters of a document. (<?xml ?> + <?xml-stylesheet ?>
         # shouldn't be that long)
-        Nokogiri.HTML(string, url, encoding,
+        Nokogiri.HTML4(string, url, encoding,
           options || XML::ParseOptions::DEFAULT_HTML)
       else
         Nokogiri.XML(string, url, encoding,
           options || XML::ParseOptions::DEFAULT_XML)
-      end.tap { |doc|
+      end.tap do |doc|
         yield doc if block_given?
-      }
+      end
     end
 
     ###
     # Create a new Nokogiri::XML::DocumentFragment
-    def make input = nil, opts = {}, &blk
+    def make(input = nil, opts = {}, &blk)
       if input
-        Nokogiri::HTML.fragment(input).children.first
+        Nokogiri::HTML4.fragment(input).children.first
       else
         Nokogiri(&blk)
       end
@@ -115,14 +86,15 @@ module Nokogiri
       Nokogiri(*args, &block).slop!
     end
 
+    # :nodoc:
     def install_default_aliases
       # Make sure to support some popular encoding aliases not known by
       # all iconv implementations.
       {
-        'Windows-31J' => 'CP932',	# Windows-31J is the IANA registered name of CP932.
-      }.each { |alias_name, name|
+        "Windows-31J" => "CP932",	# Windows-31J is the IANA registered name of CP932.
+      }.each do |alias_name, name|
         EncodingHandler.alias(name, alias_name) if EncodingHandler[alias_name].nil?
-      }
+      end
     end
   end
 
@@ -130,15 +102,27 @@ module Nokogiri
 end
 
 ###
-# Parser a document contained in +args+.  Nokogiri will try to guess what
-# type of document you are attempting to parse.  For more information, see
-# Nokogiri.parse
+# Parse a document contained in +args+.  Nokogiri will try to guess what type of document you are
+# attempting to parse.  For more information, see Nokogiri.parse
 #
-# To specify the type of document, use Nokogiri.XML or Nokogiri.HTML.
+# To specify the type of document, use {Nokogiri.XML}, {Nokogiri.HTML4}, or {Nokogiri.HTML5}.
 def Nokogiri(*args, &block)
-  if block_given?
-    Nokogiri::HTML::Builder.new(&block).doc.root
+  if block
+    Nokogiri::HTML4::Builder.new(&block).doc.root
   else
     Nokogiri.parse(*args)
   end
 end
+
+require_relative "nokogiri/version"
+require_relative "nokogiri/class_resolver"
+require_relative "nokogiri/syntax_error"
+require_relative "nokogiri/xml"
+require_relative "nokogiri/xslt"
+require_relative "nokogiri/html4"
+require_relative "nokogiri/html"
+require_relative "nokogiri/decorators/slop"
+require_relative "nokogiri/css"
+require_relative "nokogiri/html4/builder"
+
+require_relative "nokogiri/html5" if Nokogiri.uses_gumbo?

data/lib/xsd/xmlparser/nokogiri.rb

@@ -1,7 +1,9 @@
-require 'nokogiri'
+# frozen_string_literal: true
 
-module XSD # :nodoc:
-  module XMLParser # :nodoc:
+require "nokogiri"
+
+module XSD
+  module XMLParser
     ###
     # Nokogiri XML parser for soap4r.
     #
@@ -26,40 +28,40 @@ module XSD # :nodoc:
     class Nokogiri < XSD::XMLParser::Parser
       ###
       # Create a new XSD parser with +host+ and +opt+
-      def initialize host, opt = {}
+      def initialize(host, opt = {})
         super
-        @parser = ::Nokogiri::XML::SAX::Parser.new(self, @charset || 'UTF-8')
+        @parser = ::Nokogiri::XML::SAX::Parser.new(self, @charset || "UTF-8")
       end
 
       ###
       # Start parsing +string_or_readable+
-      def do_parse string_or_readable
+      def do_parse(string_or_readable)
         @parser.parse(string_or_readable)
       end
 
       ###
       # Handle the start_element event with +name+ and +attrs+
-      def start_element name, attrs = []
+      def start_element(name, attrs = [])
         super(name, Hash[*attrs.flatten])
       end
 
       ###
       # Handle the end_element event with +name+
-      def end_element name
+      def end_element(name)
         super
       end
 
       ###
       # Handle errors with message +msg+
-      def error msg
-        raise ParseError.new(msg)
+      def error(msg)
+        raise ParseError, msg
       end
-      alias :warning :error
+      alias_method :warning, :error
 
       ###
       # Handle cdata_blocks containing +string+
-      def cdata_block string
-        characters string
+      def cdata_block(string)
+        characters(string)
       end
 
       ###
@@ -69,16 +71,16 @@ module XSD # :nodoc:
       # +prefix+ is the namespace prefix for the element
       # +uri+ is the associated namespace URI
       # +ns+ is a hash of namespace prefix:urls associated with the element
-      def start_element_namespace name, attrs = [], prefix = nil, uri = nil, ns = []
+      def start_element_namespace(name, attrs = [], prefix = nil, uri = nil, ns = [])
         ###
         # Deal with SAX v1 interface
-        name = [prefix, name].compact.join(':')
-        attributes = ns.map { |ns_prefix,ns_uri|
-          [['xmlns', ns_prefix].compact.join(':'), ns_uri]
-        } + attrs.map { |attr|
-          [[attr.prefix, attr.localname].compact.join(':'), attr.value]
-        }.flatten
-        start_element name, attributes
+        name = [prefix, name].compact.join(":")
+        attributes = ns.map do |ns_prefix, ns_uri|
+          [["xmlns", ns_prefix].compact.join(":"), ns_uri]
+        end + attrs.map do |attr|
+          [[attr.prefix, attr.localname].compact.join(":"), attr.value]
+        end.flatten
+        start_element(name, attributes)
       end
 
       ###
@@ -86,13 +88,13 @@ module XSD # :nodoc:
       # +name+ is the element's name
       # +prefix+ is the namespace prefix associated with the element
       # +uri+ is the associated namespace URI
-      def end_element_namespace name, prefix = nil, uri = nil
+      def end_element_namespace(name, prefix = nil, uri = nil)
         ###
         # Deal with SAX v1 interface
-        end_element [prefix, name].compact.join(':')
+        end_element([prefix, name].compact.join(":"))
       end
 
-      %w{ xmldecl start_document end_document comment }.each do |name|
+      ["xmldecl", "start_document", "end_document", "comment"].each do |name|
         class_eval %{ def #{name}(*args); end }
       end
 

data/patches/libxml2/0001-Remove-script-macro-support.patch

@@ -0,0 +1,40 @@
+From 27e4aa8d885e47a296ea78d114dbbe8fc7aa3508 Mon Sep 17 00:00:00 2001
+From: Kevin Solorio <soloriok@gmail.com>
+Date: Fri, 1 Feb 2019 14:32:42 -0800
+Subject: [PATCH] Revert-support-html-h-b-7-1
+
+---
+ entities.c | 17 -----------------
+ 1 file changed, 17 deletions(-)
+
+diff --git a/entities.c b/entities.c
+index 43549bc5..82652f6d 100644
+--- a/entities.c
++++ b/entities.c
+@@ -623,23 +623,6 @@ xmlEncodeEntitiesInternal(xmlDocPtr doc, const xmlChar *input, int attr) {
+ 	    *out++ = 't';
+ 	    *out++ = ';';
+ 	} else if (*cur == '&') {
+-	    /*
+-	     * Special handling of &{...} construct from HTML 4, see
+-	     * http://www.w3.org/TR/html401/appendix/notes.html#h-B.7.1
+-	     */
+-	    if (html && attr && (cur[1] == '{') &&
+-	        (strchr((const char *) cur, '}'))) {
+-	        while (*cur != '}') {
+-		    *out++ = *cur++;
+-		    indx = out - buffer;
+-		    if (indx + 100 > buffer_size) {
+-			growBufferReentrant();
+-			out = &buffer[indx];
+-		    }
+-		}
+-		*out++ = *cur++;
+-		continue;
+-	    }
+ 	    *out++ = '&';
+ 	    *out++ = 'a';
+ 	    *out++ = 'm';
+-- 
+2.16.2
+

data/patches/libxml2/0002-Update-entities-to-remove-handling-of-ssi.patch

@@ -0,0 +1,44 @@
+From ffc08467744bd2305d41ca882c37fa30adf3a067 Mon Sep 17 00:00:00 2001
+From: Kevin Solorio <soloriok@gmail.com>
+Date: Wed, 27 Feb 2019 14:34:17 -0800
+Subject: [PATCH 2/2] update entities.c to remove handling of ssi
+
+---
+ entities.c | 21 ---------------------
+ 1 file changed, 21 deletions(-)
+
+diff --git a/entities.c b/entities.c
+index 43549bc5..5c4a2a60 100644
+--- a/entities.c
++++ b/entities.c
+@@ -592,27 +592,6 @@ xmlEncodeEntitiesInternal(xmlDocPtr doc, const xmlChar *input, int attr) {
+ 	 * By default one have to encode at least '<', '>', '"' and '&' !
+ 	 */
+ 	if (*cur == '<') {
+-	    const xmlChar *end;
+-
+-	    /*
+-	     * Special handling of server side include in HTML attributes
+-	     */
+-	    if (html && attr &&
+-	        (cur[1] == '!') && (cur[2] == '-') && (cur[3] == '-') &&
+-	        ((end = xmlStrstr(cur, BAD_CAST "-->")) != NULL)) {
+-	        while (cur != end) {
+-		    *out++ = *cur++;
+-		    indx = out - buffer;
+-		    if (indx + 100 > buffer_size) {
+-			growBufferReentrant();
+-			out = &buffer[indx];
+-		    }
+-		}
+-		*out++ = *cur++;
+-		*out++ = *cur++;
+-		*out++ = *cur++;
+-		continue;
+-	    }
+ 	    *out++ = '&';
+ 	    *out++ = 'l';
+ 	    *out++ = 't';
+-- 
+2.16.2
+

data/patches/libxml2/0003-libxml2.la-is-in-top_builddir.patch

@@ -0,0 +1,25 @@
+From 0b6ae484761fa01242fe8b67b54e3eb2d282d83d Mon Sep 17 00:00:00 2001
+From: Mike Dalessio <mike.dalessio@gmail.com>
+Date: Wed, 4 Dec 2019 08:43:51 -0500
+Subject: [PATCH] fix libxml2.la's path
+
+---
+ Makefile.in | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/Makefile.in b/Makefile.in
+index cf96d41..1372d8b 100644
+--- a/Makefile.in
++++ b/Makefile.in
+@@ -1057,7 +1057,7 @@ clean-noinstLTLIBRARIES:
+ 	  rm -f $${locs}; \
+ 	}
+ 
+-libxml2.la: $(libxml2_la_OBJECTS) $(libxml2_la_DEPENDENCIES) $(EXTRA_libxml2_la_DEPENDENCIES) 
++$(top_builddir)/libxml2.la: $(libxml2_la_OBJECTS) $(libxml2_la_DEPENDENCIES) $(EXTRA_libxml2_la_DEPENDENCIES)
+ 	$(AM_V_CCLD)$(libxml2_la_LINK) -rpath $(libdir) $(libxml2_la_OBJECTS) $(libxml2_la_LIBADD) $(LIBS)
+ 
+ testdso.la: $(testdso_la_OBJECTS) $(testdso_la_DEPENDENCIES) $(EXTRA_testdso_la_DEPENDENCIES) 
+-- 
+2.17.1
+

data/patches/libxml2/0005-avoid-isnan-isinf.patch

@@ -0,0 +1,81 @@
+This patch is a result of rake-compiler-dock using centos 7 (manylinux2014) to cross-compile.
+
+Centos, for reasons I have not been able to discern, implements `isnan` and `isinf` as a function
+and not as a macro. Debian knows how to resolve that function at dynamic-link time (despite using a
+macro at compile time), but musl-based systems (like alpine) do not. Running `nm` on nokogiri.so
+created on such a centos system shows:
+
+```
+                 U __isinf@@GLIBC_2.2.5
+                 U __isnan@@GLIBC_2.2.5
+```
+
+(see https://github.com/sparklemotion/nokogiri/pull/2142 for more info)
+
+This patch avoids using glibc's `isnan` and `isinf` calls, instead using libxml2's fallback
+implementation. There's history here, see libxml2 commit 8813f39:
+
+    commit 8813f39
+    Author: Nick Wellnhofer <wellnhofer@aevum.de>
+    Date:   2017-09-21 00:11:26 +0200
+    
+        Simplify XPath NaN, inf and -0 handling
+        
+        Use C99 macros NAN, INFINITY, isnan, isinf. If they're not available:
+        
+        - Assume that (0.0 / 0.0) generates a NaN and !(x == x) tests for NaN.
+        - Use C89's HUGE_VAL for INFINITY.
+        
+        Remove manual handling of NaN, infinity and negative zero in functions
+        xmlXPathValueFlipSign and xmlXPathDivValues.
+        
+        Remove xmlXPathGetSign. All the tests for negative zero can be replaced
+        with a test for negative or positive zero.
+        
+        Simplify xmlXPathRoundFunction.
+        
+        Remove Trio dependency.
+        
+        This should work on IEEE 754 compliant implementations even if the C99
+        macros aren't available, but will likely break some ancient platforms.
+        If problems arise, my plan is to port the relevant trionan.c solution
+        to xpath.c. Note that non-compliant implementations are impossible
+        to fully support, anyway, since XPath requires IEEE 754.
+
+This patch would be unnecessary if any of the following was true:
+
+* centos implements these as macros, and doesn't generate an unresolved symbol for either in the shared library
+* we had a way to ensure `__isinf` and `__isnan` resolve on musl (e.g., we implement them locally)
+
+diff --git a/xpath.c b/xpath.c
+index 9f64ab9..5b6d999 100644
+--- a/xpath.c
++++ b/xpath.c
+@@ -515,11 +515,7 @@ xmlXPathInit(void) {
+  */
+ int
+ xmlXPathIsNaN(double val) {
+-#ifdef isnan
+-    return isnan(val);
+-#else
+     return !(val == val);
+-#endif
+ }
+ 
+ /**
+@@ -530,15 +530,11 @@ xmlXPathIsNaN(double val) {
+  */
+ int
+ xmlXPathIsInf(double val) {
+-#ifdef isinf
+-    return isinf(val) ? (val > 0 ? 1 : -1) : 0;
+-#else
+     if (val >= xmlXPathPINF)
+         return 1;
+     if (val <= -xmlXPathPINF)
+         return -1;
+     return 0;
+-#endif
+ }
+ 
+ #endif /* SCHEMAS or XPATH */

data/patches/libxml2/0009-allow-wildcard-namespaces.patch

@@ -0,0 +1,77 @@
+From 74c95ec5932c737d4fcb06b8646b0017364ada14 Mon Sep 17 00:00:00 2001
+From: Mike Dalessio <mike.dalessio@gmail.com>
+Date: Fri, 24 Dec 2021 19:08:01 -0500
+Subject: [PATCH] attempt to hack in wildcard namespaces to xpath
+
+I'm not confident this is a bulletproof patch.
+---
+ xpath.c | 24 ++++++++++++++++++------
+ 1 file changed, 18 insertions(+), 6 deletions(-)
+
+diff --git a/xpath.c b/xpath.c
+index 1aa2f1a..c7f0885 100644
+--- a/xpath.c
++++ b/xpath.c
+@@ -146,6 +146,9 @@
+ #define XPATH_MAX_RECURSION_DEPTH 5000
+ #endif
+ 
++#define WILDCARD_PREFIX "*"
++#define IS_WILDCARD_PREFIX(p) xmlStrEqual((xmlChar*)WILDCARD_PREFIX, p)
++
+ /*
+  * TODO:
+  * There are a few spots where some tests are done which depend upon ascii
+@@ -11073,12 +11076,15 @@ xmlXPathCompNodeTest(xmlXPathParserContextPtr ctxt, xmlXPathTestVal *test,
+     SKIP_BLANKS;
+ 
+     if ((name == NULL) && (CUR == '*')) {
+-	/*
+-	 * All elements
+-	 */
+ 	NEXT;
+-	*test = NODE_TEST_ALL;
+-	return(NULL);
++	if (CUR != ':') {
++	    /*
++	     * All elements
++	     */
++	    *test = NODE_TEST_ALL;
++	    return(NULL);
++	}
++	name = xmlCharStrdup(WILDCARD_PREFIX);
+     }
+ 
+     if (name == NULL)
+@@ -11327,6 +11333,10 @@ xmlXPathCompStep(xmlXPathParserContextPtr ctxt) {
+ 	}
+ #endif
+ 	if (CUR == '*') {
++	    if (NXT(1) == ':') {
++		NEXT;
++		name = xmlCharStrdup(WILDCARD_PREFIX);
++	    }
+ 	    axis = AXIS_CHILD;
+ 	} else {
+ 	    if (name == NULL)
+@@ -12030,7 +12040,7 @@ xmlXPathNodeCollectAndTest(xmlXPathParserContextPtr ctxt,
+     /*
+     * Setup namespaces.
+     */
+-    if (prefix != NULL) {
++    if (prefix != NULL && !IS_WILDCARD_PREFIX(prefix)) {
+         URI = xmlXPathNsLookup(xpctxt, prefix);
+         if (URI == NULL) {
+ 	    xmlXPathReleaseObject(xpctxt, obj);
+@@ -12369,6 +12379,8 @@ xmlXPathNodeCollectAndTest(xmlXPathParserContextPtr ctxt,
+ 				    {
+ 					XP_TEST_HIT
+                                     }
++                                } else if (IS_WILDCARD_PREFIX(prefix)) {
++                                    XP_TEST_HIT
+                                 } else {
+                                     if ((cur->ns != NULL) &&
+                                         (xmlStrEqual(URI, cur->ns->href)))
+-- 
+2.31.0
+